EMS/CMP NOTES
Spring 1995
Commentary
Brett Terry, Editor
And so it is that we come to the end of another productive and exciting semester here at the Experimental Music Studios. Attempting to follow on the heels of the success of the last issue will be difficult, but this issue, once again has a compendium of valuable information to share with all the users of the Experimental Music Studios, and all those who might profit from the information we all have to share.
The only changes introduced in this issue are a consolidation of the CMP Systems Report from last issue into a more general EMS News section, which includes words from Scott Wyatt, the Director of the Experimental Music Studios, and a description of the new equipment that the studios have acquired over the past semester. The Announcement section has been reduced in size, and the Competitions section has been omitted, mostly because the sources of this information is becoming more ubiquitous in its availability on the Internet.
Since this issue also marks my departure from the Experimental Music Studios, and as editor as this publication, I would like to invite any interested parties to speak to Scott Wyatt during the summer or next Fall, about becoming the next editor, so that the EMS Notes may continue. I would also like to thank Laura Stanfield for her help in editing this issue.
User reports
James Beauchamp
Prof. James Beauchamp, comanager of CMP, has been on sabbatical leave at Paris, France since July, 1995 and is returning to Urbana in May, 1995. He has been working at the Institute for Research and Coordination in Music and Acoustics (IRCAM) principally with their group for perception and cognition of music headed by Stephen McAdams, who also heads a similar lab at University Rene Descartes (Paris V).
This year Beauchamp and McAdams have launched a project to investigate the effect of simplifying the sound spectra of a group of orchestral instruments in various ways. Frequency data are simplified by eliminating inharmonicity, rapid microvariations, and any deviations whatsoever. In a similar fashion, amplitude data are simplified by smoothing away microvariations, by forcing all harmonics to follow an identical time envelope (albeit with different scaling), and by eliminating irregularities in the spectral envelopes. Combinations of these operations can take a musical sound all the way from its original full expression to something sounding like a "buzz". The sound analysis, simplifications, and resynthesis were accomplished using Beauchamp's Unix sound analysis package, sndan. Two questions that this study will attempt to answer are: "How audible are these operations?" and "Do these operations impair listeners' abilities to distinguish sounds from those that are similar?" From informal listening experiments, it appears that this is highly dependent on the instrument being tested. Formal tests will be run this summer, and McAdams and Beauchamp expect to submit a paper on the subject early in the fall. It is hoped that the results of these tests will shed light on questions such as "What are the most important perceptual parameters for timbre?" and "How can we control them?".
In the meantime, IRCAM researchers Rene' Causse, Re'gis Msallam, and Beauchamp have begun a project to verify measurements that Beauchamp made on a trombone at UIUC several years ago. Those measurements were to obtain the frequency response (filter function) of the trombone from mouthpiece input to bell output by driving the mouthpiece with a sine wave and then to compare this the same measurement obtained using performer input. Such a filter could be (and was) used for simulating a brass instrument on a computer. However, Beauchamp found that the filter function depended on the level (dynamic) of performance, and at the time there was no available physical explanation for this phenomenon. Recently, A. Hirschberg and others at the Eindoven University of Technology made measurements that show this effect to be caused by shock waves. Their theory explains why trombone sound becomes extremely bright when the instrument is played loud. The current measurements at IRCAM will take Hirschberg's measurements further by showing in more detail the relationship between the sine wave measurements and the measurements made under performance conditions. Beauchamp serves as a consultant to this project.
Also in the meantime, Beauchamp has continued to collaborate (by email and ftp) with Andrew Horner, who is a 1993 alumnus, lecturer at Hong Kong University of Science and Technology, and cohost (with his wife, Lydia Ayers, alumnus 1994) of the 1996 ICMC in Hong Kong, on analysis/synthesis projects. During the past year they have completed projects on minimizing peak amplitudes of complex waveforms, optimizing the use of piecewise linear approximations to additive synthesis envelopes, dynamic spectral envelope synthesis of trumpet tones using a fixed wavetable and dynamic filter, and they have submitted papers on these projects to the Journal of the Audio Engineering Society and the Computer Music Journal. Beauchamp also gave a paper at the 1995 Convention of the Audio Engineering Society in Paris at the end of February on "Wavetable Interpolation Synthesis Based on Time-Variant Spectral Analysis of Musical Sounds" (AES Preprint No. 3960), with Horner as co-author. They are still working on the dynamic spectral envelope method and expect to have results on extending it to instruments besides the trumpet soon. The principal advantages of the method are that the results are very convincing, it is very efficient, and the control parameters for the synthesis are perceptually intuitive. A computer synthesis instrument which uses the method has been coded for Music 4C, and has been used, for example, in "Du Temps Perdu" by Donnacha Dennehy.
Prof. Beauchamp has been invited to give a special plenary talk at the 32nd Czech Conference on Acoustics in Prague during September 23-26, 1995. His topic will be "New Methods for Computer Analysis and Synthesis of Musical Sounds". He has received an award from the UIUC Scholar's Travel Fund to make the trip. He will also will be attending and delivering a paper on "Musical Sounds, Data Reduction, and Perceptual Control Parameters" (Andrew Horner and Stephen McAdams, coauthors) at the Society for Music Perception and Cognition Conference '95, held June 22-25, 1995 at University of California, Berkeley.
Chin-Chin Chen
This semester I have learning how to use the new equipment appropriately and efficiently. particularly the Ensoniq SQ synthesizers in Studio C. Through experiments with the new synthesizer, a variety of compositional ideas have occurred for realizing the commissioned piece within the Studio. In Studio B, I have done a 3-minute piece which was completely realized from the Moog and edited through digital devices--the Sound Designer II program, which I found to be an efficient means for editing sounds and placing sonic events in the piece. Setting music and sound effects to an about 5-minute children's story has been my other project for this semester, which is also a final project of one-year study for Electronic Music Techniques. I use concrète techniques, such as tape echo, reversal, speed change ... and generate sounds from the waveform generators and the Moog.
Donnacha Dennehy
I recently completed "GUBU (with reel climax included)" which received its first performance in the Student Works Concert of May 1. Most of the work was done in CMP, but I also made use of facilities in Studio A and in CAMIL. I recorded some sacred and profane sounds in Studio A , and the Soundhack program did the timescaling. Of course, the Midiverb proved useful now and again. All the serious programming, synthesis, and score writing was undertaken in CMP. Now that GUBU is a done job, I am ready to work earnestly (so to speak) on the commission piece.
Chris Preissing
Toylet Music is a two-channel tape piece that was finished at the end of March and premiered on an April 1 dance program produced at the stock pavilion on campus and performed again on April 2 at Smith Recital Hall. This composition reflects my interest in natural, or concrète, sounds and techniques. Most sounds were recorded in the men's toilet in the basement of the graduate library (That snare drum that you heard was not a snare drum!). Additional sounds are "borrowed" from another of my compositions, Knot of Life.
What Are You Thinking?, a 17 minute composition for two-channel tape alone, is still under construction.
Jeu de l'oie (The Game of Goose) was commissioned by Scott Wyatt for the University of Illinois School of Music Centennial Celebration. Based on chance operations derived from the Game of Goose and 16th century counterpoint rules, Jeu de l'oie is constructed from 63 pitch/timbre events and Lewis Carroll's Jabberwocky text. Jeu de l'oie, or the game of goose, is an old board game which, along with the element of chance, has a definite structure: a continuous line interrupted at specific points which, throwing vectors out to other points, creates a web-like construction. The gameboard consists of a series of sixty-three numbered squares that wraps around itself in a spiral. Casting dice, players move towards the goal of sixty-three, encountering special squares along the way. These special squares, if landed upon, require special moves --stop until another goose comes to the rescue, skip forward or backward to another square, pay a penalty. These squares are called the bridge, the well, the inn, the prison, the labyrinth, goose squares, and dead goose. Goose and dead goose require continued linear motion, either toward the goal or back to the starting point. The construction formed by the bridge, well, inn, prison, and labyrinth is an intricate web in which game players may get caught, a labyrinth superimposed upon the spiral.
The composition I propose to construct for the Centennial Celebration of the School of Music will consist of the following elements: a series of sixty-three different pitch/timbre events distributed spatially and temporally around the listener in the form of a spiral; excerpts from children's stories and prelinguistic glossolalia; and the text of Lewis Carroll's Jabberwocky recited by a man who is in the last years of his life. The 63 pitch/timbre events will be created and processed with equipment in the Experimental Music Studios of the School of Music; voices will either be recorded in the same studios or, in the case of the old man, on location in Chicago (My intention is to use the voice of my uncle, an artist, who is in a retirement home there.); other sounds will originate either from my personal sound library (accumulated over years of working in the Experimental Music Studios) or be recorded on site.
Structurally the work will be propelled by the inexorable drive through the sixty-three squares --not without a few detours through the labyrinth or a stay at the inn--and will contain sound fragments of events from throughout the life of the man who speaks the Jabberwocky text. In youth, the passing of time seems as slow as it is limitless. Likewise, the passing of the first group of the sixty-three squares will reflect this. As the years pass, time seems to go by faster and faster and likewise the rate of passing of the game squares increases correspondingly. The bridge, the well, the inn, the prison, the labyrinth, goose squares and dead goose serve as structural points that interrupt the game and may serve as entrance points to the labyrinth or some other detour. Children's stories and voices will be predominate in the first part of the composition, while the Jabberwocky text and the voice of the old man dominate the texture at the end.
Synchronicity (1989: two-channel tape) was performed at the 1995 SEAMUS National Conference in Ithaca, New York. It and Conversations (1993: piano trio) were finalists for the 1995 SCI National Conference.
In March, with Is Performance Group, a non-profit company organized in Indiana in 1993, concerts were presented at Ball State University's 25th Annual Festival of New Music and at the ELCO Theatre in Elkhart, Indiana. In April, the group traveled to Cleveland for a performance at the 8th Annual Cleveland Performance Art Festival. Pieces performed include: Knot of Life (four performers, stereo tape, clarinets, percussion, video, slides, movement), ...in the pursuit of happiness (two performers, stereo tape, slides), and Susurrus (solo dancer/vocalist, stereo tape, dance/vocal chorus).
In the wake of the successful visit by the Russian electro-acoustic composers sponsored by SEAMUS, I was invited to Moscow to participate in ALTERNATIVE-95, an exchange of Russian and American composers of electro-acoustic music. Currently, the Is Performance Group and I are pursuing grants that will allow us to make a video of Knot of Life and produce a multi-art installation/environment at the Digital Computer Lab on campus. We have been invited to perform at the Indianapolis Art Center in January, 1996 and by ArtsWatch in Louisville, Kentucky. We are organizing another series of concerts that will take us around Illinois in 1996
Steve Ricks
I've been working this semester on my piece You Expect Me, revising and re-mixing. The version I had mixed at the end of last semester had a lot of sounds I felt were too cliché (sounded like video game sound effects). But removing these sounds took a lot of the "punch" out of the piece, so I'm going back to some of the processed vocal sounds and other buchla sounds to find some material that will provide the energy the piece needs.
Some thick cluster glissandos from the buchla have added some "weight" to the piece, and seem to fill in some of the silent places that previously existed. In the last newsletter I talked about some of the processing techniques I used with Sound Designer II, but perhaps I'll mention a couple more here. I have some "harmonized" bits of spoken text, which I created by saving a single word from a playlist as a soundfile. I would then save this word several times, and pitch shift each word to a different pitch level. I then mixed these several (up to 4 at a time) words together and, voilá, a harmonized word. I have found it interesting to compress spoken text and then expand it back to close to its original length. It makes the voice sound very "electronic" and seems to accent certain frequencies and syllables in unpredictable ways.
One more thing I hope to do with this piece is to create some sound events that move from "back" to "front" in a rapid way. Then, on to the next piece.
Peter Roubal
This year I have been working on my first large-scale tape piece. Entinter Mi Casa Loca (My Crazy House), it uses Spanish and English translations of an original text. The inspiration for this work came from two seemingly disparate sources. The first is Albert Camus' The Stranger, his 1946 novel about an ambitionless man drawn for no good reason into a senseless murder on an Algerian beach. The second source is the time spent in the Spanish neighborhoods on Chicago's South side; the Pilsen and Little Village areas. Hope and ambition do exist there, but senseless destruction nevertheless takes place on al almost daily basis. It is impossible to spend any time there and not be affected by those kind of surroundings. As you might guess the piece is fairly dark in character, but what else could it be? I feel it is important for artists to be socially conscious whenever possible, but I also detest pedantic social moralizing as much as the next person. It is my intention with Mi Casa Loca merely to get people to think without telling them what to think.
All the sounds except for a bass done (realized with MIDI), were created on the Buchla synthesizer and recorded to DAT with a minimum of signal processing. All the editing was done using the Sound Designer program. The done was created with an Ensoniq KS-32.
Paul Rudolph
During this semester I have been working on a piece utilizing sounds from the Buchla, as well as concrete sounds. I altered sounds ranging from low frequency percussive attacks to high frequency cluster glissandi. I altered the sounds via tape speed variation and various Sound Tools applications.
I listened to a recording in which a French composer used assorted rolling-ball sounds. Through the use of frequency changes and panning, the composer succeeded in creating the illusion that the rolling was happening in side the listener's head. I decided to try to re-create some of those types of sounds with the Buchla and concrète techniques. Marbles rolling in a glass bowl and in a coffee can worked well, and altering those sound on Sound Tools proved interesting.
One of the difficulties I have is deciding on a program for tape pieces, but as the semester progressed and I worked on mixing the background tape, I realized that the sounds were very "subway-esque." Around this time the subway disaster took place in Tokyo, and I subsequently decided on my program: to try to depict what the person who caused the disaster might be sensing before, during and after releasing the nerve gas, i.e., what was going through the person's mind and what they were seeing and hearing.
I am content with the piece, but at my recital the sound system lacked power and depth, and I felt that the intensity was lost.
Also, the ADAT proved to be very beneficial in terms of mixing primary tapes together with ease, and allowing for immediate critique for altering phrasing, timing, etc...
Brett Terry
This semester began with the joint recital of Donnacha Dennehy, James Bohn, and myself, at which I had the following pieces performed: Nisi Dominus [1994], and Estuans Interius [1989] for female chorus, Now It is Clear [1994] for two sopranos and piano, and Mosaic [1994], for tape.
I delivered two conference papers this year, "A Object-Oriented Analysis & Design Approach to Computer Music Synthesis" at the 1995 SEAMUS Conference in Ithaca, and "CASPAR--Aural Skills Software for the Refinement of Dictation Strategies" at the 1995 Technological Directions in Music Education Conference in San Antonio. My tape piece, Mosaic, was performed at both the SEAMUS Conference and at the Biennial Arts & Technology Conference at Connecticut College. Also, my paper "ScoreViews--a Graphical, Object-Oriented Approach to Computer Music Synthesis" was accepted for the 1995 ICMC Conference in Banff.
I have spent most of the semester working on my School of Music Centennial Commission, Clotho. The meaning of the title is from Plato's Republic, wherein, there is discussion of temporality as motion through the cosmic sphere, circular time moving around the axis of linear time. Clotho, can be translated at "present" (as opposed to Lachesis, `past'), but it carries more connotations of a balance between linear and cyclic time. Although the tape part is complete, I have been recording a piano and a female chorus to achieve a rendering of the overall effect of the work for tape, amplified piano, and chorus. The piece is based on 4 original sentences. In the beginning, phonemes are associated with particular pitches, then syllables with gestures (spatial and melodic), and this type of associativity and expansion continues into the second movement, during which time there are a few complete phrase solo passages.
Andrew Walters
This last semester I finished several remixes of The Political Landscape, a piece for two-channel tape, in Studio A. Also I have been working on designing instruments in Studio C and CMP using m4c. This is in preparation for a piece to be completed this summer for flute and tape.
Articles
The ILLIAC
(Part 2 of a series)
James Bohn
In January 1961, job #1869 was submitted to the University of Illinois Digital Computer Lab by Lejaren Hiller. This job proposed the use of the ILLIAC I computer for use in the transcription of music notation. This project also used a machine that was called the Musicwriter. This machine was originally a Remington-Rand "Synchrotape" electric typewriter equipped with a paper tape unit. Cecil Effinger and Robert M. Oliver rebuilt it, changing the typeface to musical symbols and disengaging the automatic forward spacing. A printing pointer indicator was added, the backspace and shift functions were added into the control circuitry, and the platen was reduced in size. The mechanism controlling the vertical motion of the platen was altered in order to provide five functions: vertical up step, vertical downstep, vertical return to control point, clutch lock, and clutch release.
After these modifications, the Musicwriter not only functioned as a electric music typewriter, but it was also able to punch and receive five-hole paper tape. This aspect allowed it to be used in conjunction with the ILLIAC I. After a page of a score had been created with the Musicwriter, the paper tape that corresponded to the keystrokes and the platen motion could be loaded into the ILLIAC I. The computer was able to perform three functions: horizontal spacing errors, margin changes, and part extraction. Future plans included the transcription of older notation, transposition, piano reductions, and sound synthesis.
When typing out a page of score on the Musicwriter, a set of commas occurring at the left margin articulated the beginning of a system. The information occurring between two instances of this code is what was taken into account for each individual right margin 'justification.' A single comma occurring one space to the right of the left margin indicated a break between two parts within a system. Such commas were used to help coordinate the input of the information into the ILLIAC I, and were not output by the computer.
The first page of a score must be preceded with desired page margin information. This information is preceded by binary 11 (01011). The margin information (LM, RM, TM, BM) was presented with decimal numbers.
Each individual margin is terminated by binary 14 (01110). A "Print Restore" code (10110) follows this information in order to indicate that all the symbols that follow represent musical symbols. This code was also used to indicate a new page that occurred after a "Print Discontinue" code (11011) which was used to terminate the previous page. The "stop" code was ignored by the ILLIAC I, except when it followed a "Print Discontinue" code, indicating that the final page of music has been finished.
Symbols were stored in the magnetic drum memory of the ILLIAC I in a manner that reflected their location on the page. There were 9840 memory locations set aside for the storage of symbols. Each memory location was divided into 5 parts, each of which was eight bits long, yielding 49,200 possible positions per page. Upon input into the computer, each symbol was assigned an X,Y coordinate, with the upper left hand corner of the margined page equaling 0,0. This coordinate was used to assign a memory location for the character on the magnetic drum. The X value was divided by five, yielding a quotient (Q) and a remainder (R). The address that the symbol was stored at was equivalent to 2960 + Q * (Length of Page) + Y. The leftmost bit that represents the image was stored 8*R bits from the leftmost bit of the address.
The symbols themselves were represented in memory by eight bit codes. The leftmost bit was used to indicate whether the symbol was multiple or a single symbol. Multiple symbols were those that were comprised of two or three parts that are printed from the same pointer position, such as a sixteenth note, which is comprised of a notehead and a stem. In multiple symbols, the remaining seven bits correspond to a memory address that corresponds to the position within a table representing the 96 different possible multiple symbols. For single symbols, the second bit represented the shift code (upper or lower case), the third bit represented the precedence code (system A or system B), and the remaining bits represented the keystroke itself.
Once all the systems that occurred within the margins were input into the computer, the information was 'justified' by spacing it out on the drum so that the end of all the braces line up with the right margin. The output tape was then created by the ILLIAC I by first punching all the characters (left to right) that occur in the top space of the marginate page. Once this is accomplished, the characters for the next lower horizontal space is punched in the same manner, and this process is continued, until the right-most character bordering the bottom margin is output. Each horizontal space corresponds to the area taken up be one instrumental part. The end of each horizontal space was articulated by a "stop" code.
This articulation of the end of each line of a part was what allows part extraction to occur. Part extraction was done in three steps. The first step entailed the editing of the output tape of the ILLIAC I for the entire score so that only information pertaining to the desired part remains. This tape was then input into the ILLIAC I so that it could be compressed horizontally. In the memory of the computer, the information was one long line of music, so information that needs to occur at the beginning of each line (e.g. clefs and key signatures) had to be added manually after this step. Hiller had hoped to eliminate the need for editing by improving the program, but the problem was never solved. After this was done, the resulting tape could be input into the ILLIAC I for final 'justification.'
Composer Jan Bach, who was at that time a graduate student in composition at the University of Illinois, was assigned a half-time assistantship to prepare scores for publication with the Musicwriter. Unfortunately, since the ILLIAC I was removed from service late in December of 1962, very few scores were produced using the Musicwriter / ILLIAC I combination. There had been plans to rewrite the editing program for the ILLIAC III, the IBM 7090 or the CDC 1604, but none of these proposals were carried out.
Making m4c Score Files with MAX
Andrew Walters
MAX is the highly successful object-oriented program from IRCAM, well known for its ability in live interactive performance situations. Though I am interested in this approach using MAX (I should at this point refer people to the work of my colleague John Miles who has been working in this area for many years), at the present time I am using MAX to transfer material made with MIDI to m4c score files.
My interest in this stems from two concerns. First, I work in many different places on the fifth floor of the Music Building and any chance to integrate my activities is always welcome. With MAX I can bring together the work I am doing in Studio C and CAMIL with that of CMP. Along with this is the ability to mix some of the programs I work with in CAMIL (such as Finale and Performer) with m4c. Second, I am not yet the programmer I would like to be. MAX gives me a way to write "I cards" and a chance to work within a new system that will hopefully change the way I approach computer music. I am in no way stating that using MAX should replace the need and desire to program, but that it gives you a different perspective from which to work, allows you to use MIDI, and does some tasks quicker and easier (though with some tasks this is definitely not the case). The examples that I will present deal with the problem of turning MIDI into a m4c score file. There are several approaches to writing "I cards" in m4c. You could calculate all aspects of each I card one at a time, write a C program to do it for you, or use Notepro. All of these methods have their benefits and limits. My hope is to add to this the ability to create a m4c score file just by playing it back through Performer, mailing it to CMP, pasting it into a document, compiling it, and playing it back. This article is a brief overview and for this reason I will assume some basic knowledge of MAX, and will not describe specific MAX objects in great detail. I have just begun to use MAX so these results are very preliminary and any feedback would be both welcomed and appreciated.
I have chosen the Pluck instrument in the m4c class orchestra because it is has very few score parameters and is easier to understand how the components of the MAX patch and how they work. Below is the patch for changing MIDI input into an m4c Pluck "I Card "(Fig. 1).
This patch uses a Pack object to group together a number of different values and symbols into a single entity that is sent to the Text object. Each time a packed message is sent to the text object it also is sends "cr" to the Text object which tells the text object to do a carriage return. Each input has a default value (printed in the Pack object) that sends out unchanging symbols (e.g. parenthesis) and changeable values (i.e. different pitches). At the top of the Pack object are inputs for each value or signal (usually called a bang in MAX terminology). This Pack object takes in "bangs" for symbols such as parenthesis and numbers for various parameters of the "I Card" in the inputs on the top of the object.
As was mentioned before, the Pluck instrument does not contain a large number of parameters. Thus, it is possible to easily keep track of objects and their connections This particular Pack object has nine inputs for the nine values and symbols needed for an m4c Pluck "I card". MAX reads from right to left so starting on the right (input nine from the left) we have a "bang" that triggers the printing of a " }" . Since the parenthesis is already supplied in the object itself there is no need to supply it again. This is true of inputs four and five (from the right. After this a random number is supplied for the decay of the note. MIDI note and velocity information given by the Note In object is passed through an object entitled m4cconv and converted into octave-point pitch class notation and amplitude information. Start time information is
Fig. 1
supplied by various objects in the left hand side of the patch. When the first note is played the bang from the pitch information starts the Clocker object. This information is changed into seconds and sent to the Float object. When the Float object receives a note on
signal in its left input it sends the value it receives in its right input to the Pack object as a start time value. Clocker information, start time values, and note off information are sent to the m4cdur object which produces duration values. This value also triggers the Pack object to send its information to the Text object (MAX objects are trigger by input into their left-most input). This is done because the duration is the last parameter to be figured and so after each parameter's value has been received it is sent to the Text object.
Once information has been received by the Text object, by double-clicking on this object a Text window appears.
Fig. 2
I(Pluck 0.000000 0.300000 ) { 8.000000 12450 91 }
I(Pluck 0.700000 0.268000 ) { 8.030000 13050 51 }
I(Pluck 1.468000 0.228000 ) { 8.030000 15600 40 }
I(Pluck 1.796000 0.145000 ) { 8.050000 12750 60 }
I(Pluck 2.241000 0.215000 ) { 8.050000 12300 78 }
I(Pluck 2.856000 0.288000 ) { 8.080000 18300 93 }
I(Pluck 3.144000 0.125000 ) { 8.070000 9900 86 }
I(Pluck 3.269000 0.183000 ) { 8.080000 17850 86 }
I(Pluck 3.452000 0.176000 ) { 8.070000 10650 67 }
I(Pluck 3.628000 0.136000 ) { 8.080000 16050 75 }
I(Pluck 3.764000 0.161000 ) { 8.070000 10200 58 }
I(Pluck 4.025000 0.156000 ) { 8.030000 18750 38 }
I(Pluck 4.381000 0.267000 ) { 8.030000 19050 35 }
I(Pluck 4.748000 0.168000 ) { 8.050000 11700 20 }
I(Pluck 5.116000 0.261000 ) { 8.050000 18000 82 }
You may have to do some editing; it may not look as good as this. The next step is to copy and paste it into a document that can be e-mailed to CMP (or perhaps refer to Brett Terry's sheet on how to ftp files to CMP from CAMIL). [ Editor's note : This can be viewed electronically at URL http://camil45.music.uiuc.edu/SFtransfers.html ]
Then, once in CMP copy and paste it into a score file and be sure to provide an "End" statement at the bottom of the score file followed by several carriage returns. Then it can be complied like any other m4c score file.
The m4cconv2 object ( Fig. 3 ) is one I invented to transform MIDI note and velocity information into m4c score file parameters.
The input on the left takes in MIDI pitch information and splits into its pitch class and octave values. The objects on the left produce the octave information by
Fig. 3
dividing the MIDI pitch number by twelve and adding three octaves and the center objects produce the pitch classes by using modular arithmetic. It then adds these two values together into one value and sends it out the left output. The objects on the right multiply MIDI velocities by 150 and send them out the right output. I used the number 150 because if there is overlapping of notes the summation of their amplitudes will produce clipping. 251 is probably the highest number you would want to use (32000/127 = 251.97).
The m4cdur object (Fig. 4 ) receives the start time of the last "note on" value in the right input. The current time is received in the next to the right input and stored in a Float object. When the Float object receives a "bang" from a incoming "note off" value it sends the current time to an Subtraction object. Here the start time is subtracted from the current time and a duration in seconds is supplied and sent out its only output.
fig 4
There are many peculiar and less than elegant aspects to this system. Be sure before you begin to zero out the numbers associated with the start time and the numbers involved with the duration (done by supplying a "bang" to the left input of the m4cdur object). After this you can clear the Text object by sending it a clear statement (done by clicking on the box with the word clear on it). Be sure to do this in this order. One inconvenience is that the "I cards" produced by this set up produce floats with six decimal places, which can become a nuisance with more complex m4c instruments.
My initial results are preliminary (there may be timing problems with this MAX patch) in that I have worked exclusively with monophonic lines. The possibility of having a quick and efficient way to transcribe Finale files into m4c code is a appealing one. There are also many other ways to work with MIDI and m4c as well. Many thanks go out to David Berk, John Miles, and David Ralley for their invaluable help in this area. If anyone has any questions, concerns, answers, comments, condolences, etc. feel free to talk to me or drop me a line at walters@cmp-nxt1.music.uiuc.edu. I will put any new objects and patches in the MAX 2.52 folder on usr2 volume of the violetta fileserver in CAMIL.
Encapsulating Stochastic Nodes
Within a Connectionist Model
for Melody Creation
Brett Terry
Dept. of Computer Science
Univ. of Illinois
Urbana, IL 61801
bterry@uxa.cso.uiuc.edu
ABSTRACT
This paper describes a software program, Signature, that runs as a Hypercard stack on a Macintosh computer. Signature creates melodies using a set of nodes, each of which contains a set of probabilistic data of certain aspects of a set of melodies garnered through an analysis phase of the program and a weight. Through a training process incorporating both internal and user feedback, the weights of each node at each note position of a 16-note trial melody are adjusted. Signature is meant to serve as an example of a particular paradigm of creative musical knowledge, using a small data set of information at a microstructual (notes only) level. A generalization of some of the ideas presented herein might lead to a more complex version of the program on a more powerful computer, utilizing more nodes at different levels.
Analysis Material
The melodies for the analysis portion of Signature were taken from A Child's Book of Songs, by Robert Foresman. An exact listing of the melodies used is to be found in Appendix I. These melodies were, for the most part, major and diatonic. Most of the melodies were in duple (2/4 or 4/4) meter. None of the melodies used included pickup (partial) measures, since Signature's beat analysis subsystem assumes the melody begins on beat 1 of measure 1. Attempting to make the computer infer the presence of a pickup measure is a difficult task, since it must look for a note of significant harmonic importance (i.e. either scale degree 1 or 5, in most but not all cases) occurring on beat 1 of measure 1 and, upon finding such an occurrence, extrapolate to establish a best guess for the beat of the first note.
Signature is subject to some restrictions concerning the melodies it can accept. Velocity values are parsed into 8 subdivisions corresponding to pp, p, mp, mf, f, and ff. At present, only duration values between a 16th note and a half note are recognized, so no durations were used that were longer than a half note. Additionally, Signature can not recognize slurs
Since most of the melodies used did not include dynamic markings, these values were put into the melody data by the author in an effort to approximate the dynamic variations that might occur in performance (i.e. softer repeats of repeated passages, louder high notes, exclamations, and refrains), In Signature, the user is required to input the relative major pitch and the meter, pieces of information that would require a sophisticated and complex algorithm in order for the computer to deduce correctly. Connectionist models have been created that can establish a reasonable guess at the harmonic implications of a melody, such as (Scarborough, Miller & Jones 1989), (Leman, 1991), (Cope, 1991) and (Bharucha & Todd, 1991).
Given a set of melodies in standard midi file format (which can be played into a sequencer, or played out from a notation-program), Signature tabulates information about the pitch, harmonic/chordal/functional content, durations, and velocity (loudness) and stores the accumulated results. The following information is stored:
Name: first Description: probability of
Note each scale-degree occurrence in
the total number of notes
firstDisplace probability of displacing a
pitch-class one of two octaves
in either direction
stepRatio ratio of steps (movement by
minor or major second) vs. skips
secondNote probability of a scale-degree
given a preceding scale-degree
(2nd order Markov distribution)
firstChord probability of a chord's
occurrence
firstFunc probability of a function's
occurrence
secondChord 2nd order Markov chord
distribution
secondFunc 2nd order Markov function
distribution
pitchBeat probability of a specific
scale-degree given either a
"strong" or "weak" beat
chordBeat probability of a specific chord
given either a "strong" or
"weak" beat
funcBeat probability of a specific
function given either a
"strong" or "weak" beat
firstVel probability of a velocity's
occurrence
avgVel the average velocity
velBeat probability of a specific
function given either a
"strong" or "weak" beat
firstDur probability of a durations's
occurrence
avgDur the average duration
durBeat probability of a specific
duration given either a
"strong" or "weak" beat
secondDur 2nd order Markov duration
distribution
Chord and function probability distributions are calculated in the following way:
For each beat (which depends on the meter input), Signature tabulates all the notes that occur within that beat. For each note, Signature contributes one vote each to each chord associated with a particular scale degree (e.g. scale degree 1 contributes to chords I, IV, and vi). If there is only one note (e.g. a quarter note or more), or there is a clear and unique maximum chord candidate (who has received the most votes), one vote is given to that chord. Otherwise, Signature divides up the one vote among all the chords who have the maximum, and gives them each a partial vote.
Reading in the Melodies
When beginning a new set of statistics, the user can click on the "Clear accumulated Statistics" button of the "Analysis" card. Statistics about the melodies are stored within the stack, so they will be saved with the stack. To work with another set of melodies, one could use a copy of the stack, clear it, and load in new data. Weights, unlike analysis statistics, are stored in memory and must be saved and loaded from disk, if one wishes to save them. When opening the stack for the first time, the user is asked whether or not to reset the weights. This sets all weights to +1.0, in preparation for the start of a training session.
Clicking on "Analyse Midi files!" brings up a prompt dialog asking the user to specify a number of melodies to analyse, and, for each, an open-file dialog box asking for the name & location of each standard midi file. I have included these files on the attached disk. Analysis results are appended to the given statistics, unless reset with "Clear Accumulated Statistics".
The Navigator palette can be used to examine the results of the analysis. It can be selected from the Palettes menu:
In this particular study, I have chosen 18 children's melodies. A sample screen describing some analyzed characteristics are shown below:
The 1st order card includes the distributions of scale degrees, velocities, durations, octave displacements, steps vs. skips, and the average velocity and duration. For the melodies studied, notice the prominence of scale degrees 5 and 1, followed by 3, 2, 4 and 6, and 7. Velocities are weighted towards mf and mp with a few softer and louder passages. Displacements are weighted towards "0" --and the dominance of "-1" over "1" can be explained by the use of an ascending pitch numbering scheme for indicating the key, i.e. higher "key" notes, such as B and Bb in the middle of the treble clef, are more likely to have notes below them.
The 2nd order note and duration distribution card for this melody set indicates that:
* scale degree 1 is most often followed by scale degree 1, then 7 (leading-tone resolution) and 5.
* scale degree 7, which we would expect to be most often followed by 1, is equally as likely to be followed by scale degrees 6 (downward step) and 2 (cambiata figure resolving to 1).
* scale degree 2 is most often followed by scale degree 1, 2, or 3 (stepwise motion or possibly an ascending sequence of something previously in the tonic)
* scale degree 3 is most often followed by 2, then 3, 1, and 4.
* scale degree 4 is most often followed by 4 or 3, then 5.
* scale degree 5 is most often followed by 5, then 4, 6, 3, & 4.
* scale degree 6 is most often followed by 5, then 6.
The beat correlation information for pitch, velocity, and durations indicates that scale degree 1 is more likely to occur on a weak beat, 2 on a strong beat, 7 on a strong beat, 5 on a strong beat, and the rest about evenly distributed.
The beat correlation information for chords and functions indicates that most chords and functions occur with a similar distribution between strong and weak beats.
The chord and function 1st and 2nd-order distribution data indicates that tonic (I) chord harmonies are the most prevalent, followed by vi, iii, IV, and V. Alternate functions are more prevalent than Tonic because this functionality subsumes two chords, whereas the Tonic function includes only one chord (I).
The absence of tonic-dominant movement is a characteristic of this type of children's melody, i.e. when there is a break from the stepwise motion it is usually a skip to the tonic, not from a tonic-harmony note to a dominant-harmony note.
The 2nd order distributions indicate that:
* I chords are most commonly followed by iii chords, then IV..
* ii chords are most commonly followed by IV chords, then iii (see biasing explanation above).
* iii chords are most commonly followed by V chords, then vi (biasing).
* IV chords are most commonly followed by vi chords, then V and IV.
* V chords are most commonly followed by I chords.
* vi chords are most commonly followed by I chords (biasing).
* viideg. chords are most commonly followed by I chords.
* tonic chords are most commonly followed by alternate chords, then sub-dominant
* sub-dominant chords are most commonly followed by alternate chords, then more sub-dominant chords (prolongation) and dominant chords.
* dominant chords are most commonly followed by tonic chords
* alternate chords are most commonly followed by tonic chords, then dominant chords.
Training
The process of training involves both self-reflexive feedback (using the rules described) and user feedback. For the purposes of training, all generated melodies are in the key of C and in 4/4 time, consisting of the first 16 notes.
At the start of each melody creation, the delta values (combined user feedback and internal rules) are added to the current weights for each note position, i.e. a separate weight exists at each position in the 16-note sequence for each node.
User Feedback
User feedback works in the following way. If a pitch or duration is marked "don't like", Signature finds all nodes who suggested that choice and subtracts .33 to their deltaWeight (which is applied prior to next melody creation). If an alternative is suggested (this can be left blank), Signature looks for all nodes who suggested that note or duration choice and adds .33 to their deltaWeight at that note position. Since the chord and function nodes suggest chord and function values rather than note choices, Signature tests to see if the user's alternative is to be found in any of the other chord or function choices and adds .2 to their deltaWeight if so. Note that one ramification of this procedure is that if the user selects an alternative that was not suggested by any node (and usually there are only three or four), it has no effect. If a displacement is marked "Don't Displace" then the chance of any displacement occurring at all at that note position is lowered by .33.
It is hoped that this type of feedback would be less of a drawback in a larger system where a greater number of contributing nodes would increase the number of distinct candidates. Although this system does provide an effective method of feedback, it is undeniably true that the user's alternative will never be a "correct" response. The "goal" of feedback thus does not, and should not, aspire to attain to any objective criteria for "correctness" but rather, the goal is to become closer to the subject's musical aesthetic in this particular setting. The name "Signature" arises from this idea of a subjective stylistic "signature" circumscribed by a larger horizon of historical harmonic practices.
A "Play Mine" button is provided that will play the version of the melody substituting user choices for pitches and durations, and canceling displacements. In order to register the feedback, the user must click on the "Feedback" button.
Internal Rules
So far, only the note, chord, and function weights have non-random internal rules. These are the stepRule, and the consensusRule. There is also a randomRule, in an attempt an impetus for evolution through variation.
The stepRule votes for either a step or a skip, based on the step/skip ratio. Then it multiplies the current stepRatioWeight at that note position by -.1 or .1 to inhibit or excite any of the 4 note producing nodes which suggested either skips or steps.
The consensusRule compares the (weighted) results of the 4 note-producing nodes, the 3-chord producing nodes, and the 3-function producing nodes. It calculates the disparity of each group, i.e. how many different votes for note, chord, and function, and then maps this onto a deltaWeight (e.g. a disparity of "1" = .67 (all agree), down to .1 (all disagree)). If the chord, function, and note choice are all commensurate, each node in each group who voted for the final candidates are rewarded by the sum total of the other two disparity-delta-weights. Thus, we all get rewarded if we all agree, and rewarded even more if one or more of us was really sure about the candidate they were suggesting. If two nodal-types agree, they are each rewarded and the dissenter is penalized to an equal degree. If all three disagree, they are all penalized.
The randomRule just picks up to 10 weights (out of a 17 weights x 16 position weight space) and adds a random number between -.05 and +.05 to the chosen weights.
Expectations
One motivation behind creating Signature is see if, over the course of time, more detailed analyses (such as the 2nd order functions) will prevail over less detailed (1st order functions) which, in turn, will prevail over random functions.
Working as a composer suggests to me, that, in the act of musical creation, one draws on a wealth of diverse information relevant to the harmonic practices of Western European Art music. The model of musical creative intelligence am interested in investigating has as a basic the notion that there are dynamic balances being struck between composing "by-the-rules" and decisions that appear to be made intuitively, yet are also made, in some larger sense, for larger "aesthetic" reasons of contextual musical dramaturgy. My hope is that enough tests of Signature, even with such a small set of analysis nodes, might provide a suggestive basis for the belief that, given a much larger set of nodes (perhaps implemented on a supercomputer), a long-enough training period might provocatively hint at unpredictable combinations of dominating weight values corresponding to some goal-oriented feedback system. Some other issues regarding refinement have been discussed in (Lewis, 1991).
Although it would be extremely useful to have demonstrative evidence of the success of the feedback process, such a result would most likely only be evident after a substantial (between 100 and 10000) number of feedback sessions. Given the amount of time involved in such a trial, that will remain an ongoing procedure, whose outcome will eventually be known.
Ideas for the Future
Another level of analysis that I began designing was a pattern-level analysis (in contradistinction to the note-level analysis used so far). One significant drawback of the note-level probabilistic analysis methods I have used so far is their inability to discern repetitions and patterns within a melody. Although perhaps computationally infeasible (at least within Hypercard), one idea I developed was similar to the Auto-Correlation method used for pitch detection. For this method, a copy of the melody is slid horizontally, beat-by-beat against itself and scale-degree differences are computed. A local maxima in the difference function may imply a sequence. This is, however, complicated by the fact that differences have to computed as tonal transpositions (the half-step between scale degrees 4 and 3 should be akin to the whole-step between scale degrees 5 and 4) rather than chromatic transpositions (exact numbers of half-steps).
Another possibility is the incorporation of previously research models for quantization of musical time (Dessain 1989), patterns (Gierdingen 1989), and multiple constraints (Mozer, 1991).
Conclusion
When one composes a piece of music, one is constrained, in a certain sense, not only by note-level choices, but by pattern-level and gestalt-level (the work as a whole) choices, as well as temporally-regressive choices (the more one has already written, the more it, to some degree, will affect the remainder of what is still to be written). Applying quantitative methodologies to these complex interactions remains difficult, but especially so if one is driven towards singular, unified theories. One advantage of this system described herein is that it allows for a pluralistic set of analysis sets and a process of evolution whereby the dynamic balance between sets (nodes) can be modified, through feedback, to approach more closely a subject's contextualized aesthetic. Any sense of progress can thus reveal exactly the balances used, and, in a generalized more complex example, this type of revelation might prove ultimately informational in characterizing aspects of a signature stylistic preference.
Bibliography
Bharucha, Jamshed, and Peter Todd [1989], "Modeling the Perception of Tonal Structure with Neural Nets" in Music and Connectionism, ed. Todd & Loy, Boston : MIT Press, pp. 3 - 15. Reprint ed from Computer Music Journal 13(4).
Cope, David [1991]. Computers and Musical Style, Madison : A-R Editions, Inc,
Desain, Peter, Henkjan Honing, and Klaus de Rijk [1989], "The Quantization of Musical Time: A Connectionist Approach" in Music and Connectionism, ed. Todd & Loy, Boston : MIT Press, pp. 3 - 15. Reprint ed from Computer Music Journal 13(3).
Dolson, Mark [1989] "Machine Tongues XII: Neural Networks" in Music and Connectionism, ed. Todd & Loy, Boston : MIT Press, pp. 3 - 15. Reprint ed from Computer Music Journal 13(3).
Gierdingen, Robert [1989] "Using Connectionist Models to Explore Complex Musical Patterns" in Music and Connectionism, ed. Todd & Loy, Boston : MIT Press, pp. 138 - 159. Reprint ed from Computer Music Journal 13(3).
Leman, Marc [1991] "The Ontogenesis of Tonal Semantics: Results of a Computer Study" inMusic and Connectionism, ed. Todd & Loy, Boston : MIT Press, pp. 100 - 126.
Lewis, J. P. [1991] "Creation by Refinement and the Problem of Algorithmic Music Composition" in Music and Connectionism, ed. Todd & Loy, Boston : MIT Press, pp.212 -228.
Loy, D. Gareth [1991] "Connectionism and Musiconomy" in Music and Connectionism, ed. Todd & Loy, Boston : MIT Press, pp. 20 - 37.
Mozer, Michael [1991] "Connectionist Music Composition Based on Melodic, Stylistic, and Psychophysical Constraints" in Music and Connectionism, ed. Todd & Loy, Boston : MIT Press, pp. 191 - 211.
Olson, Harry F. [1967], Music, Physics, and Engineering (2nd ed.), New York : Dover, 1967, pp. 430 - 444.
Scarborough, Don, Ben Miller, and Jacqueline Jones [1989] "Connectionist Models for Tonal Analysis" in Music and Connectionism, ed. Todd & Loy, Boston : MIT Press, pp. 54 -61 Reprint ed from Computer Music Journal 13(3).
Todd, Peter [1989] "A Connectionist Approach to Algorithmic Composition" in Music and Connectionism, ed. Todd & Loy, Boston : MIT Press, pp. 173 - 194. Reprint ed from Computer Music Journal 13(3).
Appendix I
List of Melodies used
Taken from A Child's Book of Songs, by Robert Foresman
1. Down the Lane (p.5)
2. Columbine (p. 8)
3. Halloween (p. 9 - Russian)
4. Come And Wander (p. 13 - Frence)
5. The Clock (p. 15 - German)
6. The School Bell (p. 6)
7. Little Things
8. Train Song (p. 16)
9. Questions (p, 17 - Brandenburg)
10. Little Birdie (p. 18 - A.E. Johnstone)
11. On The Winter Wind (p. 22 - Czech)
12. A Game of Tag (p. 23 - Old Southern Melody)
13. How d'ye do my Partner (p. 25 - American)
14. The Hippopotamus and I (p. 26 - Hungarian)
15. Careless Paul (p. 28 - Norwedian)
16. The Proper Kitten (p. 29 - Italian)
17. Morning Hymn of Praise (p. 33 - German)
18. The Elves Dance at the Wedding (p. 35 - French)
19. Running Races (p. 36 - Polish)
20. The Piper (p. 56 - English)
Reducing Noise of the QuadraVerb
Peter Roubal
Although I've come to appreciate the sounds and versatility of the new Alesis QuadraVerb, I found, frequently that the unit produced a substantial amount of noise hiss. One technique I found effective for removing some of the hiss was a combination of the Quadraverb's internal EQ and the comb filters on the Buchla.
Enhanced Spatial Panning Using the Alesis ADAT and HyperPrism
Brett Terry
With the arrival of the new ADAT and HyperPrism, I found some interesting tricks for creating interesting panning effects in studio A. Typically, I would begin with a mono file in Sound Designer, and paste two copies of it into a stereo file, as the left and right tracks. Then I would apply a "fade in" and "fade out", one to each track. This I would record to two tracks of the ADAT, in effect, having a sound source that traveled from 1 track to the other (albeit linearly). Next, I would user the mixer and set a particular stereo placement, reverb, and delay setting for the output of each ADAT track, to create the effect of motion from one definite location to another. Other enhancements to this technique include using HyperPrism's HyperStereo or MoreStereo effects in thru mode, or the Quadraverb's auto-panning effect.
Automated Editing with the Buchla
Richard Power
One of the compositional concerns in my piece Zoometry is the abrupt, unprepared changes from one environment to another. This is easily achieved with basic editing techniques, but I soon realized that this procedure could quickly become stale. I then began to wonder how I might halt only certain layers of an environment while leaving the others to continue into the next. The difficulty came in the fact that I wanted each edit to precisely coincide with a percussive sound, so the impression would be that the changes in activity were being caused by these other sounds.
A solution to the problem was found by using the envelope follower and sequencer on the Buchla. I had three layers of material: one stereo and two mono. These I fed into the Buchla's four VCA's (with levels set at 0), and then to the Yamaha mixer. I sent the recording of my percussive sounds to the mixer and also to the input of one of the Buchla's envelope followers.
I used each of the four TV outputs of the 16-stage sequencer to trigger the VCA's, arranging the output levels so that they would enable either all three layers, each of the three combinations of only two layers, or only one of the layers to be triggered. The DC output of the envelope follower was then used to advance the sequencer one stage for every pulse.
Inspired by these results I decided to set up a patch in which a single environment would be triggered through varying equalization settings by a short sound. These settings would be pronouncedly different from each other, but the activity within the environment would be such that the listener would be able to hear its continuity. Cutting and splicing techniques would result in an obvious discontinuity between sound envelopes. (In some cases this could be remedied through careful matching of the envelopes, but this was not the situation here because of the density and speed of my environment. More importantly, automating the process saves a great deal of time.)
The patch was similar to the one above except that a single mono source was sent to each of the four VCA's. Each VCA output was then sent to a mixer input where they were equalized. The short sounds triggered the sequencer again via the envelope follower, but this time only a single VCA was triggered at a time.
While I was very happy with the results of this type of procedure, it obviously could be developed further. Using the 4-stage sequencer and the Roland's VCA's would enable one to use many more layers. Also, by varying the TV voltage amounts at each stage of the sequencer the user could change amplitude from one stage to the next, perhaps creating an environment in which the layers' spatial relationships changed abruptly and at a fast speed. The possibilities are many, and I suggest this method of automated control as a way to exploit compositional ideas which are only possible within the electro- acoustic domain.
Helpful hints for the new studio user
Chin-Chin Chen
The following, probably known by every current user, could be useful hints for the new learner and I would like to share my experience with them.
1) When one gets a new blank 1/4" tape "pancake", there is a header stuck at the beginning of the tape to avoid the tape coming out of the "pie". Don't throw that away after you get the tape onto the reel. That tiny thing is very useful for keeping the tape on the "pancake". Simply stick the header back on to the beginning of the tape after you have done your job and you can store your "pancake" securely without worrying that the tape will come loose.
2) The leader tape used in Studio B is a 1/4" plastic timing leader tape. Every two segments equal to the length of one second at the speed of 15 ips. When you need to put 6 feet (72") of leader tape both at the beginning and the end of your piece, simply cut a length of 10 segments of the leader tape and those equal to approximately 75". By so doing, you don't need to measure the tape with a ruler.
3) I still remember facing a pile of spliced tapes lying on the floor, and I have to be very careful so that I won't step on them accidentally (believe me that it is a nightmare if it happens). I found the rack for the leader tape serves a very good spliced-tape collector. Simply take the leader tape off the rack and put your empty reel onto it. Now you can manually get the tapes onto the reel immediately after you splice them together. Of course, remember to put the leader tape back to the rack after you finish your job.
4) A check list here may be useful for troubleshooting for whoever uses the program DATa to backup/restore their sound files:
* Did you turn DigiDesign on?
* Is the channel 1, 2 input set to "Digital" in Hardware Setup menu item under the Setup menu (for restore) and is the input light on (which means set to Digital) on DAT tape deck two (for backup)?
* Did you put the DAT tape to the right deck? (to deck one for restore, to deck two for backup)
* Did you close the write protect window on the DAT tape so that the tape is writeable, when you backup your file?
While you backup your sound file, always check if there is any signal coming out of the DAT deck level meters so that you won't lose your file. DATa restores your sound file by reading the header and then the data onto DAT tape when you backup the file. Always wait until the data is written as well, then push the stop button on the DAT deck when you backup your file. I had a data file without a completely written amount of post-header data and couldn't restore it. I tried to rewind the DAT tape to the beginning of the file and let DATa read the header of the same file; luckily I was later able to restore that file.
5) If you encounter with memory capacity problems while creating a Sound Designer file. You would probably like to check:
* Is the trash can empty? If not, empty it. Always open the trash can and make sure you only trash what is supposed to be thrown away.
* If you are working in Studio E, make sure that you save your file onto External Hard Drive. When you create a new sound file or when you try to restore your data file, a window will pop up asking you to name the file. Select the desktop until the EXT icon shows up in the upper right corner, then write the filename, and hit return. Don't confuse yourself and the computer!
6) A few features in Sound Designer II, I think, are very useful as well. There are two ways of editing in Sound Designer II. One is non-destructive and the other, destructive. I use destructive editing mostly. Under the DSP menu, there is Parametric EQ and Graphic EQ. They both are very useful for getting rid of segmented hiss or noise. I personally prefer Parametric EQ. The Dynamics command can increase or decrease the amplitude of your sound source. "Pitch shift" and "Time Compression/Expansion" are also useful; however, according to my experiments, I feel that they work well only with a certain kinds of sounds. Some sounds will get distorted after making changes through those two devices. How much change you make with your original sound also needs to be taken into consideration.
One effect I like the most is to reverse only one channel of a stereo sound, and shift both channels a certain amount of time (it depends on what kind of sound you have). Thus produces a quasi panning effect on the condition that the original sound has fade-in or fade-out quality (instantaneous attack and decay sound won't work).
In the soundfile window, the Overview Display shows the entire soundfile, either as a waveform or a time line. If you would like to quickly locate different sound events in the piece or you would like to see the entire soundfile while listening to the piece, it can be a better to set the overview display show a waveform line instead of a time line.
If you are using destructive editing and need to synchronize sound events in two files, you would probably like to set the Time Axis to display Hr:Min:sec:Msec, which we use in everyday life, instead of showing decimal minutes. Pay attention to the Data Indicator boxes, especially the upper two boxes. The upper left-hand box displays the distance from the insertion point or selector start to the mouse's current position. By looking at that you will know how much material you need to cut or paste to get two events happening at the same time, or if the sound you have is too long or too short to fill out the time you planned. The upper right-hand box gives you the absolute position of your mouse pointer or associated cursor relative to the beginning of the file. By looking at that, you know where to insert the sound events. These two small boxes help you to have more control of synchronization.
7) Last but not least, be patient while working with the equipment. They take time to respond your requests If you rush and keep pushing buttons or hitting keys on the keyboard, you may confuse the machines and they may finally "quit their job". As we all know that we can not afford to lose our studio time while trying to meet the deadline, we are all responsible for maintaining the equipment in their optimum condition.
EMS News
Scott Wyatt, Director of the Experimental Music Studios
Centennial Commissions
During the Fall semester, I announced the creation of the University of Illinois School of Music Centennial Celebration [1895-1995] Electro-Acoustic Music Commissions The submission process was open to all students working within the School of Music Experimental Music Studios (CMP being part of EMS).
Works were to be composed and realized specifically for this commission, and must be realized within the Experimental Music Studios' facilities. Length of each work are to be between 6 and 9 minutes (no longer than 9 minutes). Proposed works could be for tape alone or for instrument(s) and/or voices with tape.
Composers interested in being considered for a Centennial Commission were asked to submit a one page proposal describing the project in as much detail as possible, along with a tape of two representative works. Finalists were determined based upon the quality and feasibility of the proposed work and the quality of the representative works submitted. Submissions were reviewed by a selected group of UIUC doctoral composition students who have completed their prelims and who have worked within the Experimental Music Studios (comprised of David Berk, Judd Danby, and Richard Power).
Awarded commissions were announced on February 17, 1995. Commissions were awarded to:
James Bohn
Chin-Chin Chen
Donnacha Dennehy
Christopher Preissing
Brett Terry
Andrew Walters
Christopher Weise
Congratulations to these composers!
New Hardware and Software Installed
This semester we were successful with the installation of the following:
CMP Hardware:
Tascam-DA30 Mk II DAT deck
HP LaserJet 4 600 DPI LaserPrinter
CMP Software:
Christopher Penrose's command-line sound processing utilities (/usr/local/bin/sound3)
(type: "fputils" to see current list)
/SoundApps/Csnd 1.6 - Stephen David Beck, Csound front-end
/SoundApps/cmixEZ.app 1.0 - Paul Lansky, Cmix front-end
/carl/cmusic
/SoundApps/InstrumentBuilder .99 - Mark Trombino (UCSD), cmusic front-end
/SoundApps/ReFill.app - Arun Chandra
/LocalApps/SynthBuilder (demo)
/usr/local/bin/X11/mosaic - mosaic client for X/co-Xist
/usr/local/bin/X11/mxv - X-based sound editor by Doug Scott
/usr/local/bin/sound2 CARL command-line function and sound processing utilties ("moresndutils")
/usr/CMP/instruments - Added csound & cmix instrument files with sample score files
Studio A: Hyperprism audio processing software, additional memory for the Quadra 650, Alesis ADAT, Alesis Quadraverb,
Studio B: Yamaha REV-7
Studio C: Ensoniq SQ-2 synthesizer, Ensoniq SQR rackmount synthesizer, Sony A7 DAT recorder, Alesis Midiverb III
Studio D: Kyma System, additional memory for the Centris 650, memory DAT tape drive, Alesis ADAT, Alesis Midiverb II
Studio E: Mac IIcx with DigiDesign Soundtools software, additional memory for the Mac IIcx, memory DAT tape drive, external 500 MB hard disk, DATa and Retrospect backup software
Yet to be installed is:
Studio D: Kurzweil K2000 synthesizer
Some EMS Farewells
We have had an active and productive year within the studios' facilities, and I would like to congratulate all users for their efforts, successes, and contributions. Several users are leaving the University at the end of this semester and are moving on to other horizons, jobs, and more degree work. I would like to extend my appreciation and best wishes for future success to:
Paul Anderson, Russ Breimeier, John Miles,
Paul Rudolph, and Seth Wrightington.
We are also saying farewell to five assistants who have worked as EMS/CMP staff over the past year or more. Their contributions have been very significant to the operation, instruction, composition and research interests occurring within the facilities. Best wishes and good luck to:
Tim Dixon, Cheri Herndon, Richard Power,
David Ralley, and Brett Terry.
Thank you for all of your assistance. We wish you the very best in your continuing careers.
Announcements
DIFFUSION i MeDIA wishes to announce the on-line verson of their catalog with Candadian Electro-Acoustic music :
Jean-Francois Denus
empreintes DIGITALes - SONARt
4487, rue Adam / Montreal, QC H1V 1T9 / Canada
T +1 514 254-7794 / F +1 514 281-1884
URL http://www.cam.org/~dim/index.html
The Formation of the Electronic Music Foundation
The formation of the Electronic Music Foundation came about in September of 1994 with the goals being to create a photo archive of the history of electronic music, a newsletter containing information regarding the location and availability of historical materials and important events and projects, and the worldwide sale of compact discs of electronic music. Joel Chadabe is the president of EMF with Paul Lansky, Neil Rolnick, and Joel Chadabe serving as Board of trustees. The foundation has formed an international group of advisors and charter subscribers that included Jon Appleton, Larry Austin, Marc Battier, John Bischoff, Jack Body, Tom Buckner, Warren Burt, Joseph Celli, Chris Chafe, Hugh Davies, John Duesenberry, Folkmar Hein, Gottfried-Michael Koenig, Paul Lansky, Otto Laske, George Logemann, Peter Manning, Max Mathews, Eduardo Miranda, Stephen Montague, Dexter Morrill, Lawrence Moss, Randy Neal, Pauline Oliveros, Gottfried Raes, Tom Rhea, Alistair Riddell, Jean-Claude Risset, Curtis Roads, Neil Rolnick, Carla Scaletti, Wayne Siegel, Bruno Spoerri, Carl Stone, Robert Scott Thompson, Kazuo Uehara, Alvise Vidolin, Felix Visser, Patte Wood, Scott Wyatt, and Iannis Xenakis.
eMUSIC[TM] Launched
As part of the goals of the Electronic Music Foundation, eMUSIC has been created. Serious music lovers worldwide will now have access to any and all compact discs of experimental, exceptional, and/or electronic music -- including hard-to-find CDs, and discs published by small companies or independent composers. The recordings are being marketed via international computer networks and direct mail.
Joel Chadabe, President of EMF, said: "Our goal is to find and contact everyone in the world who has an interest in experimental and electronic music, and to make CDs available to them through mailed and electronic catalogs."
For more information about eMUSIC and its catalog of musical offerings, contact Electronic Music Foundation via telephone at (518) 434-4110 (voice), via fax to (518) 434-0308, or send an email message to: eMusc@aol.com. Or write to: Electronic Music Foundation, 116 North Lake Avenue, Albany NY 12206, USA.
WELCOME TO eMUSIC!
=============================== APRIL 1995
eMUSIC is a new worldwide service that -- for the first time -- gives you easy, direct-mail access to any and all compact discs of experimental, exceptional, and/or electronic music. Wherever you live, eMUSIC brings you CDs that may be hard to find, discs published by small companies or independent composers, even recordings you may not have known existed.
If you're a composer, compact discs of your works will be available through eMUSIC. If you're a listener, eMUSIC gives you immediate access to some of the world's most interesting music. If you're both -- well, better yet!
How does it work? Every month or so we'll send you a brief catalog of selected discs available through eMUSIC. You'll find electronic music history, unusual sounds, new ways of combining words and music, new approaches to improvisation, computer music, interactive music, portraits of composers, 20th-century virtuosi...and more! Our goal is to make available every disc of experimental, exceptional, and electronic music -- in short, eMUSIC.
In this introductory catalog, we're setting the tone (so to speak). You'll find compact discs from the United States, Canada, France, Australia, Germany, and Switzerland. You'll find two unusual packages representing the two first European studios of the 1950s: a definitive collection of Pierre Schaeffer's work in musique concrete and a collection of Stockhausen's early work in Cologne. And you'll find a special offer.
But if you don't find what you're looking for, just ask for it. For information on how to order CDs, read the "How to Order" section that follows the listing of discs. We suggest you also read the section called "Join the Inner Circle."
=============================== THE DISCS
PIONEERS: and a special! Take 'The Art of the Theremin' free with a Charter Subscription!
===============================
THE ART OF THE THEREMIN
First demonstrated in 1920, the Theremin is played by moving one's hands in the air. Clara Rockmore is its first virtuoso, accompanied here by Nadia Reisenberg, pianist. The package includes an informative booklet. It's all beautiful magic.
=> eMUSIC #DE-100 ($15, or free this month with Charter Subscription)
PIERRE SCHAEFFER: MUSICAL WORKS
The definitive collection, including 'Etude aux Chemins de Fer' ('Railroad Study', 1948), the first piece of musique concrete, and 'Symphonie pour un Homme Seul' (1950, 'Symphony for One Man Alone'). Four discs and a book of essays (in French) and photographs.
=> eMUSIC #IN-106 $49
STOCKHAUSEN: ELEKTRONISHE MUSIK 1952-1960
Includes 'Etude' (1952), composed at Pierre Schaeffer's studio in Paris, 'Studie I' (1953), 'Studie II' (1954), 'Gesang der Junglinge' (1956), and 'Kontakte' (1960). Also a book of essays and articles on the music in German and English, scores, and photographs.
=> eMUSIC #ST-100 $36
ELECTRO ACOUSTIC MUSIC CLASSICS
Includes Edgard Varese' 'Poeme Electronique' (1957), first performed at the Philips Pavilion at the Brussels World Fair in 1958 in a building designed by Iannis Xenakis who was at the time working for Le Corbusier as an architect; Iannis Xenakis' 'Mycenae-Alpha' (1978), composed with Xenakis' original UPIC System, a graphic input device that let him "draw" sounds; Milton Babbitt's 'Philomel' (1963) with sounds of the RCA Mark II Sound Synthesizer and Judith Bettina, soprano; and Roger Reynolds' 'Transfigured Wind IV' (1985), computer sounds and flutist Harvey Sollberger.
=> eMUSIC #NE-103 $16
ROARATORIO
One of John Cage's big big big works, composed in 1979, with Cage reading his text 'Writing for the Second Time Through Finnegans Wake', Irish musicians playing and singing, and all the sounds mentioned in Finnegans Wake as assembled by Cage on a 62-track collage tape. Two discs include a conversation between Cage and Klaus Schoning (who commissioned the work for the WDR in Cologne) and Cage reading his text. Also two booklets with essays, conversations, and the text.
=> eMUSIC #MO-108 $26
FORBIDDEN PLANET
The original 1956 soundtrack to MGM's 'Forbidden Planet' by Louis and Bebe Barron. In their words, "We created individual cybernetic circuits for particular themes and leit motifs..." One of the first electronic film tracks.
=> eMUSIC #GN-100 $15
JAMES TENNEY: SELECTED WORKS 1961-1969
Tenney began at Bell Labs at the dawn of computer music. He led the way. This disc includes 'Collage #1 ("Blue Suede")' (1961), 'Analog #1: Noise Study' (1961), and other early pieces.
=> eMUSIC #AR-106 $14
RISSET
Classics of elegant computer music. Includes 'Inharmonique' (1977), which relates sound and harmony; 'Sud' (1985), based on sounds of the sea near Marseilles; 'Dialogues' (1975), for flute, clarinet, piano, and percussion with taped computer sounds; and 'Mutations' (1969), composed at Bell Labs in New Jersey.
=> eMUSIC #IN-103 $16
JOHN CHOWNING
Breakthrough moments in computer music, including 'Sabelithe' (1966, revised 1971), 'Turenas' (1972), 'Stria' (1977), and 'Phone' (1981). Chowning aimed at sounds that were special to computers yet expressive and beautiful. In 'Stria', for example, the sounds are simply out of this world.
=> eMUSIC #WE-102 $19
MORTON SUBOTNICK: TOUCH
Subotnick was there to play the first Buchla synthesizer in 1963. In 'Touch' (1969), he touches things to control the sound. 'Jacob's Room' (1986), on the other hand, is hi-tech drama with Joan La Barbara, soprano. From 'Touch' to 'Jacob's Room' is technology time travel.
=> eMUSIC #WE-104 $19
TERRY RILEY: IN C
A fabulous performance of music that changed the world. It's about patterns, shifting tonalities, energy, all around C. 30 musicians playing 30 instruments plus. 'In C' was originally commissioned in 1964 by the San Francisco Tape Music Center.
=> eMusic #NA-120 $16
===============================
COMBINING WORDS AND MUSIC:
in opera, sound poetry, and improvisation
===============================
TRANSPARENCE
Marc Battier, in Paris, transforms -- using Phonogramme, a graphic audio program -- a fragment of speech by Henri Chopin, a well-known French pioneer of sound poetry. You hear his voice for a few seconds only, but it's under everything. As Battier puts it, "The voice of the poet... is reincarnated as a new skin, it changes form, it is in mutation, it is re-invented." It also produces extraordinary sounds.
=> eMUSIC #UN-100 $16
KURT SCHWITTERS: URSONATE
'Ursonate', a verbal sonata, is a legend of sound poetry, finished in 1932. Invented words are used as motives, rhythms and themes, and combined in the form of a sonata. Eberhard Blum brings more than twenty years of performances to this disc.
=> eMUSIC #HA-216 $19
eL/AFICIONADO
An electronically-accompanied opera by Robert Ashley, in his words "scenes from the life of an 'agent'... 'debriefing' to a jury of Interrogators..." With Thomas Buckner, Robert Ashley, Jacqueline Humbert, and Sam Ashley, all singing and speaking. The textures and electronic orchestrations are, as in all of Ashley's music, seductive, beautiful and evocative. Libretto included.
=> eMUSIC #LO-101 $16
73 POEMS
Joan La Barbara's vocal-electronic settings of Kenneth Goldsmith's multi-level, evocative, sometimes lyrical, sometimes funny, 73-part poem. It's an exceptional marriage of words and word-sounds as music.
=> eMUSIC #LO-123 $15
MACHINE FOR MAKING SENSE
Five from Australia. Chris Mann and Amanda Stewart read and sing their poetry while Stevie Wishart plays violin and hurdy-gurdy, Jim Denley plays winds, and Rik Rue manipulates sounds, and everyone plays with electronics. Quite an incredible, rich, complex, and wonderful improvisation.
=> eMUSIC #OO-117 $15
VALIS
An electronic opera by Tod Machover, commissioned by the Pompidou Center in Paris for its 10th anniversary in 1987. It's based on the science fiction novel by Philip K. Dick. Many voices sing and speak with keyboard, percussion and electronic techniques developed at IRCAM and MIT. Original English version. Libretto included.
=> eMUSIC #BR-101 $16
===============================
PLAYING TOGETHER: unusual approaches to group activity.
===============================
THE HUB: COMPUTER NETWORK MUSIC
Music by John Bischoff, Tim Perkis, Chris Brown, Scot Gresham-Lancaster, Mark Trayle, and Phil Stone, all in the San Francisco area. Their "Hub" is a central computer which takes what someone plays and interprets it for the others so that they're all acting on the same data. It's a network approach to group performance.
=> eMUSIC #AR-101 $14
CONCERTO GROSSO
Richard Teitelbaum composed this -- well, he devised the system that produced it -- in 1985. Improvisations by Teitelbaum (keyboard and computer), Anthony Braxton (saxophones), and George Lewis (trombone) alternate with computer-controlled acoustic grand pianos and synthesizers. Lively and musical.
=> eMUSIC #HA-118 $19
NO WORLD (TRIO) IMPROVISATIONS
Joseph Celli and Jin Hi Kim, with Alvin Curran, Malcolm Goldstein, Shelley Hirsch, Adam Plack, and Mor Thiam, come together as a global-village improvisation group. There's an Australian didjeridoo, Korean komungo and piri, African percussion, oboe, English horn, violin, voice, and electronics, and sounds like never before.
=> eMUSIC #OO-102 $15
MOVIN' ON
Updated jazz. Bruno Spoerri plays saxophones, Synthophone (electronic saxophone invented by Martin Hurni in Switzerland), synthesizers, and a Macintosh computer. Reto Weber does a lot with percussion. With Albert Mengelsdorff, trombonist, and Ernst Reijseger, cellist.
=> eMUSIC #TU-100 $15
THOUSAND YEAR DREAMING
Parts of it fully scored, parts improvised, Annea Lockwood has assembled Jon Gibson, Art Baron, Libby Van Cleve, J.D. Parran, Michael Pugliese, Scott Robinson, John Snyder, Charles Wood, and Peter Zummo to play didjeridoo, conch shell, trombone, oboe, English horn, clarinet, contrabass clarinet, tam-tam, clapping sticks, frame drums, pod rattle, tam-tam, superball, waterphone, and stones. The title refers to Australian aboriginals' sense of history as dreaming, and the titles of the sections-breathing and dreaming, the Chi stirs, floating in mid-air, in full bloom-describe the moods.
=> eMUSIC #WN-109 $15
===============================
MUSIC ABOUT SOUNDS: finding and making new sounds at home and abroad.
===============================
DENIS SMALLEY: IMPACTS INTERIEURS
For Smalley, "The starting point is always the sound..." 'Valley Flow' (1991) is the sound of the Canadian Rockies. And what a stunning sound! Also included: 'Piano Nets' (1991), 'Wind Chimes' (1987), 'Clarinet Threads' (1985), and 'Darkness After Time's Colours' (1976).
=> eMUSIC #IM-106 $15
MILA'S JOURNEY INSPIRED BY A DREAM
Eliane Radigue synthesizes atmospheres. Her sounds fill the air, suggesting peace, time and transformation. This sound is an environment for Milarepa's songs (Milarepa was an 11th-century Tibetan poet who taught through singing), sung here by Lama Kunga Rinpoche in Tibetan and recited by Robert Ashley in
English.
=> eMUSIC #LO-113 $15
HOMEBREW
Paul Lansky takes banged kitchen objects, traffic noises, fairytales, clapping, a shopping mall, and transforms the sounds. It's an homage to his family. Can computer music be charming? Yes.
=> eMUSIC #BR-107 $15
ANGELS AND INSECTS
David Dunn looks to the biosphere: "Beneath the water's surface are a variety of plants and small insects... While the sounds above water are comfortable and familiar, those occurring under the surface are shocking... as if controlled by a mysterious but urgent logic..." They make for compelling music.
=> eMUSIC #WN-108 $15
RAINFOREST IMAGES
Priscilla and Barton McLean went far afield for these sounds, among them Peruvian and Australian bird calls, wooden recorders, and didjeridoo, all processed electronically in studios in Australia and New York.
=> eMUSIC #CA-104 $15
OCEAN FLOWS
Rik Rue, Australian sound artist, records sounds and creates soundscapes. 'Ocean Flows' is about ocean waves over beaches and rocks. 'Ebbs Tides and Flows' uses the sounds of Sydney Harbor. 'Goondiwindi', an Aboriginal word meaning "water flowing over rocks," is a water-sound exploration of crevices, channels and caves. It's hearing a whole new world in what we think we know.
=> eMUSIC #TP-103 $19
===============================
COMPILATIONS: surveys, group
portraits, overviews, variety.
===============================
THE VIRTUOSO IN THE COMPUTER AGE (II)
Music from five US studios, including Larry Austin's 'Life Pulse Prelude' (1984), based on sketches for Charles Ives' unfinished Universe Symphony; Gareth Loy's 'Blood From A Stone' (1992), for Max Mathews' electronic violin; Chris Chafe's and Dexter Morrill's 'Duo Improvisation' (1991), with celletto (Chafe's electronic cello), trumpet, and computer; Neil Rolnick's 'The Persistence of the Clave' (1992); Rodney Waschka II's 'Last Night' (1990), for alto sax and piano; Jon Appleton's 'Homenaje a Milanes' (1987), for Synclavier; and Larry Polansky's '(And to rule...)(Cantillation Study #2)' (1988), for flute and computer. Assembled by Larry Austin.
=> eMUSIC #CE-111 $15
SEAMUS
A selection, mostly for taped sounds with and without acoustic-instrument performances, compiled by the Society for Electro-Acoustic Music in the United States. Includes Larry Nelson's 'Order and Alliance'(1991), Scott Wyatt's 'Counterpoints' (1992), Joseph Koykkar's 'Triple Play' (1992), Joseph Anderson's 'in mosaic' (1992), Charles Mason's 'The Blazing Macaw' (1992), Stephen Beck's 'Love's Not Time's Fool' (1992), Eric Chasalow's 'Fast Forward' (1988), and Paul Koonce's 'Whitewash' (1992).
=> eMUSIC #SE-100 $15
JEWEL BOX
Music, performance art, and radio works by Anne LeBaron, Laetitia Sonami, Sussan Deihim, Bun Ching Lam, Catherine Jauniaux, Ikue Mori, Sapphire, Mary Ellen Childs, and Michelle Kinney, all exceptional women composers and performance artists with something to say. Sometimes something spellbinding. Storytelling and sounds from harp to household objects to electronics to electronically treated speech.
=> eMUSIC #TE-101 $16
AUSTRAL VOICES
Exceptional voices from Australia. Alan Lamb records "telegraph wires singing in the wind." Alistair Riddell plays a computer-driven piano. Sarah Hopkins extends cello and vocal sounds in the "natural unhurried rhythms of the environment of the Northern Territory." Warren Burt composes for his specially made and tuned tuning forks. Ros Bandt records in a hollow concrete cylinder five floors underground in Melbourne. Jeff Pressing plays sequencers and synthesizers. Ross Bolleter uses the ruined piano at the Nallan sheep station north of Perth.
=> eMUSIC #NA-107 $16
===============================
COMPOSERS: discs that are indicators
to a single composer's approach.
===============================
FRANCOIS BAYLE: FABULAE
Bayle is director of the Groupe de Recherche Musicale in Paris and inheritor of the tradition of musique concrete. He stretches the tradition. His sounds start from acoustic activities -- xylophone glissandi, bird calls, wind, percussion, harps, water bubbling -- but the total effect is of an unlimited orchestra accompanying an imaginary ballet.
=> eMUSIC #IN-124 $16
GIUSEPPE G. ENGLERT
Englert was a founding member of the Groupe Art et Informatique de Vincennes (GAIV) in Paris. He's always looked for new ways to compose, to think. 'Sopra la Girolmeta' (1991), the electronic work here, is about algorithms and interactivity. Also: 'Les Avoines folles' (1963) for string quartet; 'Fragment & Caracol' (1974) for orchestra; 'La Joute des Lierres' (1966) for string quartet; and 'Babel' (1983) for orchestra.
=> eMUSIC #GR-101 $19
PAULINE OLIVEROS: CRONE MUSIC
Originally composed to accompany the actors' amplified dialog throughout Mabou Mines' production of 'Lear', here it's performed by Oliveros on her expanded accordion -- which employs electronics to process the instrument's sound during performance.
=> eMUSIC #LO-112 $15
JERRY HUNT: GROUND
Hunt performed with devices -- usually his own inventions, often involving visuals and collaborations with other artists -- linked together in systems. As he puts it, "congruent layers of associations of inflection-calls evoke... point specific, melody-action strings embedded in a reference context of conventions of performance." At some point in your life, you must experience the originality of Jerry Hunt's personality.
=> eMUSIC #OO-107 $15
DEXTER MORRILL: GETZ VARIATIONS
Underneath Morrill's computer music lurks the soul of a jazz musician. It's in the turns of the phrases, moods, rhythms, colors. 'Getz Variations' (1984), 'Sketches for Invisible Man' (1989), and 'Six Studies and an Improvisation' (1992) all involve saxophones and computers, the lyrical and the interesting. David Demsey plays the saxophones.
=> eMUSIC #CE-120 $15
ALCIDAS LANZA: TRILOGY
Lanza grew up in Argentina, lived in New York, and now lives in Montreal. 'Trilogy' is an autobiographical cycle of songs presented as an evening of music theatre for singer, electronic sounds, and lights, with texts in English, French, Spanish, and an invented language, sung here by the impressive Meg Sheppard.
=> eMUSIC (#SH-100) $16
ROBERT NORMANDEAU: LIEUX INOUIS / UNHEARD-OF PLACES
Normandeau's unheard-of places are in the scenes depicted by his sounds, of wind, footsteps, water, words... and synthesized electronic sounds, composed in Montreal, Quebec, Bourges (France), and Ohain (Belgium). With evocative audio imagery veering between narration and abstract sound, his music casts a spell.
=> eMUSIC #IM-101 $15
MICHEL REDOLFI: DESERT TRACKS
Redolfi directs the Centre International de Recherche Musicale in Nice, but he's also spent a bit of time in California. In 'Desert Tracks' (1988), you feel the space and heat of the Mohave, of Death Valley -- in Redolfi's words, "the shivering resonances of the landscapes..." Also: 'Too Much Sky' (1984) and 'Pacific Tubular Waves' (1979), inspired by the California coast. Redolfi's sounds glitter like the waves.
=> eMUSIC #IN-105 $16
CARL STONE: MOM'S BAR-B-Q
"Sometimes," Stone says, "I'm simply attracted to a kind of wonderful moment..." So he takes the moment and makes something special, like the extraordinary sound transformations in 'Banteay Srey' (1991) and 'Shing Kee' (1986). Also: 'Gadberry's' (1989), 'Mom's' (1990), and 'Chao Nue' (1990).
=> eMUSIC #NA-114 $16
NEIL ROLNICK: MACEDONIAN AIR DRUMMING
Rolnick gets around. He was in Yugoslavia for the sounds in 'Macedonian Air Drumming' (1990), which he performs with Air Drums. And 'ReRebong' (1989) uses Indonesian gamelan sounds. 'Sanctus' (1990), a score for Barbara Hammer's film, is electronically-processed fragments from masses by Machaut, Byrd, Bach, et al. 'Balkanization' (1988) uses Balkan folk songs.
=> eMUSIC #BR-105 $15
MARY JANE LEACH: CELESTIAL FIRES
Here's music -- 'Bruckstuck', 'Green Mountain Madrigal', 'Mountain Echoes', and 'Ariel's Song' -- performed by the New York Treble Singers, a chorus of women's voices which makes the sounds float like clouds on a breezy day. There's also 'Feu de Joie' and 'Trio for Duo', instrumental and vocal pieces.
=> eMUSIC #XI-106 $15
PHILL NIBLOCK: FOUR FULL FLUTES
Niblock records instruments, in this case played by flutists Petr Kotik and Eberhard Blum, and mixes the sounds in multiple layers and clusters to form a big sound, sometimes strident, sometimes harmonic. As Tom Johnson said of Niblock's music, "It must be heard to be believed." It's true.
=> eMUSIC #XI-100 $15
BOB WILLEY: PEACE PIECES
Willey lives in California and, in his words, this disc is "new age jazz bossa nova folk electronic instrumental experimental." Actually, it's more than that: it's California new age jazz bossa... It's also Norway, Casteneda, and other places and writers, and it puts a hi tech approach in the service of easy listening.
=> eMUSIC #OU-100 $15
===============================
HOW TO ORDER
===============================
To order discs, contact eMUSIC via email, fax, regular mail, telephone, personal visits -- or, in extreme cases, by carrier pigeon. Give us a list of the CDs you'd like to buy, and mention their eMUSIC catalog numbers. Please bear in mind that some discs are available only in relatively small numbers. We'll fill orders as we receive them.
Tell us your address (including email) and phone numbers, and where you'd like the discs shipped (if it's different from your mailing address).
Please pay in advance. You can calculate the amount of your payment as follows:
Add together the prices of the discs you'd like to buy. Then add the appropriate amount for shipping (priority mail) according to the following table:
North America & Mexico: First disc $2.00, each additional disc $1.00.
South America: First disc $3.00, each additional disc $1.50.
Europe: First disc $4.00, each additional disc $2.00.
Asia/Australasia: First disc $5.00, each additional disc $2.50.
Then, if you live in New York State, add 8% sales tax to the total. (New York State requires that we charge sales tax also on shipping charges.)
We accept credit cards (Mastercard and VISA). We also accept checks, bank drafts and money orders -- but only in US$, drawn on a US bank.
If you're paying by credit card, we need your name (as it appears on the card), your card number, the expiration date of the card, and your billing address and phone number (in case we need to contact you). Please give us credit card information by fax, phone, or regular mail (rather than email) to protect the confidentiality of your information. After you've provided us with your credit card information, you can email your orders -- and just tell us to charge your card.
Place your order by:
Email to: eMusc@aol.com
Fax to: (518) 434-0308
Telephone to: (518) 434-4110
Regular mail and carrier pigeon to: eMUSIC, 116 North Lake Avenue, Albany NY
12206, USA
===============================
JOIN THE INNER CIRCLE
===============================
eMUSIC is a program of Electronic Music Foundation (EMF), a not-for-profit organization in New York State, USA. EMF exists to disseminate information and materials related to the history and current development of electronic music.
We invite you to become a Charter Subscriber to all of EMF's activities by making a one-time gift of $100. Your financial support will help us a great deal and we'll do everything we can for you in return.
As a Charter Subscriber, you'll become a part of the eMUSIC "Inner Circle" and receive special discounts, exclusive offers, and free compact discs. If you're looking for CDs not listed among our current offerings, if you need more information, if you have discs that you'd like to bring to our attention -- in short, if you'd like to become a more integral and interactive part of the eMUSIC process -- we'll be there to respond, to help, or to work with you. You'll also receive special notifications and newsletters from Electronic Music Foundation.
If you're a professional, look at your Charter Subscription as a business investment. As a listener, think of your contribution as reserving the best seat in a "virtual concert hall."
===========================================
eMUSIC and Inner Circle are trademarks of Electronic Music Foundation, Inc.
Electronic Music Foundation
116 North Lake Avenue
Albany NY 12206
USA
Voice: (518) 434-4110
Fax: (518) 434-0308
email: eMusF@aol.com
The Experimental Music Studios
School of Music
University of Illinois at Urbana-Champaign
1114 W. Nevada St.
Urbana, Illinois 61801
EMS & CMP Director: Scott Wyatt
CMP Co-Managers: James Beauchamp, Sever Tipei
EMS Assistants:
Tim Dixon, Technical Assistant
Richard Power, Operations Assistant
CMP Assistants:
Brett Terry, Systems Assistant
David Ralley
Cheri Herndon
CMP User Advisory Committee:
James Bohn
Camille Goudesenne
Brett Terry
Ray Watts
EMS User Advisory Committe:
Richard Power
Brett Terry
Andrew Walters